While a number of HIV prevention programs have been evaluated in school settings, much of the focus of this research has been the content of the curriculum; as a result, we still know little about the specific facets of the classroom environment that may yield the strongest effects of an effective school-based FHV, other STD, and pregnancy prevention curriculum. In the proposed research, we plan to bring together expertise from our research in health communications and our research in classroom instruction, combining Donohew's individual differences model of information exposure with Eccles' stage-environment fit theory to assess the impact of curricular, teacher, instructional and student individual difference variables on HIV-related mediating and outcome variables while developing an effective prevention program. Specific aims of the project are, within a randomized control trial in high school health classrooms, to (1) adapt a modified version of a skills-based, MV, other STD, and pregnancy-prevention curriculum to be more appropriate for the needs of impulsive decision makers, in order to bring about desired behavior change in sexual decision-making; (2) assess the impact of characteristics of the instructional environment in health classrooms (i.e., teacher efficacy, teacher immediacy, value orientation, and teacher identity) in order to identify effective instructional methods and teacher characteristics; (3) implement reinforcement and relevance of the curriculum across other classes and content areas in order to foster more effective learning of the material; and 4) Evaluate the interaction among the student, the teacher, and the curriculum to determine optimal learning environments that lead to sustained behavior change. The proposed study, including over 5,000 students and 10-12 schools, will be conducted over a period of five years. During the first three years, six separate intervention implementation studies will occur (two consecutive interventions per year), with the design for five of the six intervention studies as experimental, with randomization occurring by school. In Years 4 and 5, we will also examine the effects of a combined "best practices" classroom based intervention by comparing two groups of students randomized by schools: a comparison condition, with a skills-based HIV/pregnancy prevention curriculum or the optimal version that contains the most successful elements tested in the three preceding years. The analytic design will include individual, classroom, teacher, and school levels, to be assessed through hierarchical linear modeling or other tests of differences.